Global dimming and brightening versus atmospheric column transparency in Europe, 1906–2013

 

 One of nontraditional meteorological parameters is the broadband direct solar beam irradiance (also referred to as integral direct solar beam, direct solar radiation, direct beam, etc) observed in conditions of a non-screened, clear of clouds solar disc. Information on direct irradiance is essential to assess quantitavely the variability of column transparency and turbidity and to make climatological conclusions concerning volcanic impact, urban and rural pollution, radiation regime, role of atmospheric circulation as a carrier of aerosols, conditions for remote sensing. Evaluation of possible values of direct solar beam is important in agriculture, forestry, solar energetics, architecture, material science, recreation business etc.

 

                In this work, the main parameter is the integral (broadband) transparency coefficient (p2) transformed to optical mass m = 2 (solar elevation ≈ 30°). This coeffi-cient enables easy calculation of several other broadband parameters of column transparency and turbidity (column transmittance, optical depth, the Linke turbidity factor) as well as transition to spectral Aerosol Optical thickness, at 500 nm, AOT500.

 

                We present multiannual time series of p2 for three different European climatic locations:

1)       Estonia, for two meteorological stations, Tartu-Tõravere and Tiirikoja; the first series of high quality observations of direct solar beam, during 1932-1940, was carried out using the Ångström pyrheliometer 197; after the war, the observations were started again in 1950; since 1999 the Tartu-Tõravere actinometric station has been included into the Baseline Surface Radiation Network, BSRN;

2)       Moscow (the Meteorological Observatory of the Moscow State University), 1955-2013, the observatory is situated in the southwest part of Moscow, on the territory of the large Botanical Garden of the University; the distance between the observatory and the main building of the university is about 500 m only;

3)       the southernmost site, Feodosiya is located on the eastern part of the Crimean peninsula, on the steep coast of the Black Sea, within 50 m of the coastline.

 

                For comparison, historical evolution of p2 at Pavlovsk, 1906-1936, is given. This is an historical satellite town, located 26 km south from Saint Petersburg. The town was one of the most splended residences of the Russian imperial family - this fact indirectly characterizes the air quality.

 

 Figure 1. Time series of observed direct solar beam were used from six locations: Pavlovsk, Tiirikoja, Tartu, Tõravere, Moscow, and Feodosiya; 1, 2 and 3 indicate thermal electrical power stations.

 

                These time series, covering together a 108-year period, 1906–2013, have in general a similar pattern which to some degree can be extended to the entire European latitudinal belt, 44°–60° N. At the named locations, annual means of column transparency are mainly affected by three factors:

1)       strong volcanic eruptions which can be detected, by a pyrheliometer or actinometer, up to 24 months after the eruption;

2)       human induced aerosol emissions from burning of fossil fuels which, correlating with trends in the amount of the emissions, cause multiannual smooth changes in transparency;

3)       smoke from seasonal forest and bog fires, which can be local, but also can arrive, by atmospheric circulation, from remote distances exceeding 1000 km.

 

                The highest annual mean column transparency belongs to the northernmost location Pavlovsk where p2 = 0.813 in 1909. But there were also other years with high column transparency in Pavlovsk: during the 31- year observational period the annual mean p2 = 0.8 was reached in 6 years. For further evaluations, we consider this level as an example of “very good transparency” corresponding to a year with particularly clean air.

The lowest annual value, p2 = 0.632 in 1912, belongs again to Pavlovsk and was caused by the eruption of the Katmai/Novarupta (Alaska) volcano in the same year. The Pinatubo (1991) impact was slightly less: in 1992 annual mean coefficient of column transparency decreased at two Estonian to p2 = 0.68- 0.69 and in the Crimea and Moscow to p2 = 0.65- 0.66.

 

                A series of four successive volcanic eruptions during 1979-1982 ending with the El Chichón, caused drop in transparency down to the value p2 = 0.70 in Estonia in 1983-1984, and to p2 = 0.67 in the Crimea and Moscow in 1983.

 

                Listed impact of remote volcanos is the first common feature in multiannual trends of transparency.

Second, from 1945 there is a long decreasing tendency in column transparency which lasted until 1983/1984. Lower transparency means more aerosol particles participating in cloud formation. In this way lower transparency should lead to more cloudiness and to lower global solar radiation (global = direct + diffuse). This phenomenon, known as “global dimming” indeed happened in most areas of Europe.

 

                The third common feature in multiannual pattern of column transparency is that the decreasing trends in annual mean values ended in 1983/1984. After that the trends changed their sign from negative to positive. Recovering of column transparency was interrupted by the Pinatubo eruption in 1991 which impact lasted two years. Years of improvement of transparency, during 1984 to 1991, and from 1994 onward, are in line with increase of global solar radiation, i.e. with “global brightening”.

Figure 2. Multiannual time series of column trans-parency in European locations. Black triangles at the bottom indicate dates of the largest volcanic eruptions. Triangles and circles on top indicate US-Canadian and Russian wild fires. Coefficient p2 = 0.905 would correspond to a CDA (clean and dry atmosphere).

 

Acknowledgements

The research has been supported by a project “Estonian radiation climate“ funded by European Regional Development Fund (ERDF).